摘要
利用SPOT5遥感影像数据和同期获得的野外调查样地数据,基于按植被类型分类估测的方法,研究了河北省丰宁满族自治县植被地上生物量的遥感估测技术。研究结果显示,SPOT5影像的4个波段反射率和中红外植被指数(VI3)结合建立的多元回归模型,可用于森林生物量的遥感估测,估测的R2值达0.540,说明中红外波段信息提高森林生物量的估测精度有一定作用;通过分析样地生物量与多种植被指数的相关性发现,基于比值植被指数(RVI)的指数回归模型是灌丛生物量估测的最佳模型,估测的R2值达0.711,基于归一化植被指数(NDVI)的简单线性回归模型为估测草地生物量的最佳模型,R2值达0.790。利用2008年的全覆盖SPOT5影像,获得了丰宁县2008年植被地上生物量分布图,除农田植被外,全县地上生物总量为3.706×107 t,单位面积生物量平均为51.223t/hm2,其中,森林植被总生物量为3.578×107 t,灌丛植被总生物量为1.048×106 t,草地植被总生物量为2.277×105 t。
The techniques for estimation of aboveground biomass were studied in Fengning Man Autono-mous County,Hebei Province,using SPOT5images and filed sample measurement data based on the esti-mate method by vegetation types.The study results shown that a multiple regression model,which was composed of 4band spectral reflectance of SPOT5images and Mid-infrared Vegetation Index（VI3）derived from SPOT5images,can be used for estimating forest biomass,the coefficient of determination（R2） reached 0.540.Through the correlation analysis of bush and grassland biomasses with various vegetation indices,it was found that the exponential model based on Ratio Vegetation Index（RVI）was the optimal model for estimating bush biomass,and the NDVI-based simple linear model was the optimal model for es-timating grassland biomass,with R2 values of 0.711and 0.790.The distribution map of biomass in Fengn-ing county was made using SPOT5images acquired in 2008.The map reflect that the total aboveground bi-omass（except the croplands）in 2008in Fengning is 3.706×107 t,with a average of 51.223t/hm2,in which,the total forest biomass is 3.578×107 t,the total bush biomass is 1.048×106 t,and the total grass-land biomass is 2.277×105 t.
出处
《遥感技术与应用》
CSCD
北大核心
2010年第5期639-646,共8页
Remote Sensing Technology and Application
基金
国家863计划项目(2006AA120108)资助
